How long can a losing player sun run for?

Because live poker sample sizes are so small, could there be a losing player who has a winning career because they ran like God?

Last week quite a few people shared my article about variance in High Roller tournaments. In it, we showed that given the low win rates in these tournaments (because they are all filled with crushers) and the fact they run less often, a talented player could, in theory, lose for a decade in them through no fault of their own.

That is only one side of variance, and I wanted to explore the flip side of this with a losing player and how well they could run. Once again using the tournament variance calculator at PrimeDope, we replicated the conditions in the last article. That is $30k average buy-in, 80 runners on average, 10 get paid, 6% rake and 200 MTTs played a year. The only difference was I made the player a -15% losing player, rather than a winner. 

This is what we got from the simulations, first the distributions and then 20 random samples (from 1,000 overall): 

Our player loses $900,000 on average and their chances of having a losing year are 76%. However, good variance happens. In our best run, the player won $5 million winner, making them by any standard a 'Player of the Year' contender. 

Short-term variance is well documented, so let's 10x things and make it 2,000 tournaments, to represent a decade in the game:

Our player is now down $9 million on average with one simulation seeing them lose $20 million in a decade. There is a 98% chance of loss for this player, which means that we will have a handful of outliers who still won. Our best simulation saw the player win almost $5 million over a decade.  

The big takeaway from this, for me, is that somewhere in the world there are hard-working smart players who cannot stop losing in live poker, through no fault of their own. However, there will 100% also be players who are 'losing players' in terms of their ability, but who consistently make money. In our simulation, we had a player who averaged $500k a year for a decade. There will be players in real life who have gone professional and succeeded for a long time because of their short-term luck in live poker, oblivious to the reality of their ability. 

It wasn't till I got to 4,000 MTTs, ie. 20 years, that all the simulations ended up losing. 

I also did these simulations with losing players in large field MTTs with lower buy-ins, and in theory, there would be players who went their entire career over many decades without finding out they were technically a losing player (and winners who would never make a profit). Large-field live MTT players will never see the long term. Play around with PrimeDope yourself, it's fascinating. 

This is all a stark reminder of the importance of sample size, something we perhaps take for granted in online poker where we can play a lifetime's worth of poker in a couple of months. 

Did these stats surprise you? Let us know in the comments: